Bug 20368 - VirtualBox netinstall ends with display corruptions (when switching to vboxvideo drm device)
Summary: VirtualBox netinstall ends with display corruptions (when switching to vboxvi...
Status: NEW
Alias: None
Product: Mageia
Classification: Unclassified
Component: Installer (show other bugs)
Version: Cauldron
Hardware: x86_64 Linux
: release_blocker normal
Target Milestone: Mageia 6
Assignee: Mageia tools maintainers
QA Contact:
URL:
Whiteboard:
Keywords:
: 20504 (view as bug list)
Depends on:
Blocks:
 
Reported: 2017-02-28 09:52 CET by Daniel BEZIVIN
Modified: 2017-04-26 14:28 CEST (History)
9 users (show)

See Also:
Source RPM: x11-driver-video-fbdev, kernel, virtualbox
CVE:
Status comment: Still valid in current ISOs


Attachments
Installer report.bug.xz (from upgrade) (139.30 KB, application/octet-stream)
2017-04-14 22:33 CEST, Martin Whitaker
Details

Description Daniel BEZIVIN 2017-02-28 09:52:45 CET
Description of problem:
Mageia cauldron sta2 .Net install launched yesterday evening.

I let the installation go on all the night (9 hours duration announced) .
When i come back this morning, the installation was waiting for root password but the screen was double and very small. 
However, i succeded in finishing the installation, and Mageia sta2 has rebooted perfectly with a normal display.
Comment 1 Nicolas Lécureuil 2017-02-28 09:59:14 CET
hi,

i saw this too but "not all the time" :(

thierry any idea ?


if i reproduce i will take a picture
Comment 2 Frédéric Buclin 2017-02-28 13:05:39 CET
I can reproduce with net install too, using VBox (if you needed another confirmation).
Comment 3 Barry Jackson 2017-02-28 16:09:58 CET
This looks like a duplicate of:
https://bugs.mageia.org/show_bug.cgi?id=20354
(Installer graphics die at start of users screen in VBox)
Comment 4 Marja van Waes 2017-03-01 06:35:30 CET
(In reply to Barry Jackson from comment #3)
> This looks like a duplicate of:
> https://bugs.mageia.org/show_bug.cgi?id=20354
> (Installer graphics die at start of users screen in VBox)

And there Barry reported updates (incuding kernel updates ) fixed it.

@ Daniel

Is it fixed for you, too?
Comment 5 Daniel BEZIVIN 2017-03-01 10:56:20 CET
(In reply to Marja van Waes from comment #4)
> (In reply to Barry Jackson from comment #3)
> > This looks like a duplicate of:
> > https://bugs.mageia.org/show_bug.cgi?id=20354
> > (Installer graphics die at start of users screen in VBox)
> 
> And there Barry reported updates (incuding kernel updates ) fixed it.
> 
> @ Daniel
> 
> Is it fixed for you, too?

Hi Marja
Yes the bug is fixed, no duplicated or small screen, but i encountered a new problem :
At the choice of the monitor , the system has frozen.
I wait a moment but no mouse or keyboard responding.
I took a "instantané " (Menu/ machine "host T").I stopped the machine.
Then i restored it and the installation went on without any problem.
Comment 6 Frédéric Buclin 2017-03-16 17:59:05 CET
*** Bug 20504 has been marked as a duplicate of this bug. ***
Comment 7 Thierry Vignaud 2017-03-16 20:01:38 CET
@Thomas: any idea why fbdev does this?
Comment 8 Rémi Verschelde 2017-04-04 10:44:53 CEST
Can it still be reproduced? Comments in bug 20354 (which is apparently related if not a duplicate) hint that it may be fixed now.
Comment 9 Nicolas Lécureuil 2017-04-04 11:10:20 CEST
i got it yesterday with prestesting isos
Comment 10 Rémi Verschelde 2017-04-04 11:30:38 CEST Comment hidden (off-topic)
Comment 11 Rémi Verschelde 2017-04-04 11:50:59 CEST Comment hidden (off-topic)
Comment 12 William Kenney 2017-04-04 20:37:30 CEST
This bug is still valid today.
Comment 13 Martin Whitaker 2017-04-14 22:33:41 CEST
Created attachment 9199 [details]
Installer report.bug.xz (from upgrade)

Just hit this when doing a mga5 -> mga6 upgrade in VirtualBox, using the Apr 7 Mageia-6-rc-x86_64-DVD.iso. In my case at least, the display corruption occurs when the system switches from using the framebuffer device to the vboxvideo drm device. The key section from the system log (extracted from report.bug) is:

<6>[  215.672960] [drm] Initialized
<4>[  215.688367] vboxguest: loading out-of-tree module taints kernel.
<4>[  215.691310] vgdrvHeartbeatInit: Setting up heartbeat to trigger every 2000 milliseconds
<6>[  215.691411] input: Unspecified device as /devices/pci0000:00/0000:00:04.0/input/input5
<4>[  215.691510] vboxguest: misc device minor 56, IRQ 20, I/O port d020, MMIO at 00000000f0400000 (size 0x400000)
<7>[  215.691511] vboxguest: Successfully loaded version 5.1.18 (interface 0x00010004)
<6>[  215.694294] [drm] VRAM 01000000
<6>[  215.694460] [TTM] Zone  kernel: Available graphics memory: 2024542 kiB
<6>[  215.694462] [TTM] Initializing pool allocator
<6>[  215.694466] [TTM] Initializing DMA pool allocator
<7>[  215.694648] checking generic (e0000000 1000000) vs hw (e0000000 1000000)
<6>[  215.694650] fb: switching to vboxdrmfb from VESA VGA
<6>[  215.694720] Console: switching to colour dummy device 80x25
<6>[  215.694788] fbcon: vboxdrmfb (fb0) is primary device
<6>[  215.694907] Console: switching to colour frame buffer device 100x37
<6>[  215.694920] vboxvideo 0000:00:02.0: fb0: vboxdrmfb frame buffer device
<6>[  215.699951] [drm] Initialized vboxvideo 1.0.0 20130823 for 0000:00:02.0 on minor 0

and from Xorg.0.log:

[   215.569] (II) config/udev: removing GPU device /sys/devices/pci0000:00/0000:00:02.0/drm/card0 /dev/dri/card0
[   215.569] (II) config/udev: Adding drm device (/dev/dri/card0)
[   215.569] (II) xfree86: Adding drm device (/dev/dri/card0)
[   215.574] (II) LoadModule: "modesetting"
[   215.574] (WW) Warning, couldn't open module modesetting
[   215.574] (II) UnloadModule: "modesetting"
[   215.574] (II) Unloading modesetting
[   215.574] (EE) Failed to load module "modesetting" (module does not exist, 0)
[   215.574] xf86: found device 0
Comment 14 Rémi Verschelde 2017-04-26 12:57:04 CEST
Would this issue be related to bug 20354 too?
Comment 15 Chris B 2017-04-26 14:28:55 CEST
Remi: 20354, 20368, 20504 are the same bug.

Note You need to log in before you can comment on or make changes to this bug.